I was chatting with a friend, and she mentioned how she tries to at least set up a README, which includes her vision for the project and her plan for the implementation, design, and goals.

Best case scenario is that the planning helps her complete the project herself. Worst case scenario, someone else can pick up where she left off and use her considerations for the project.

I’m thinking of doing that for future projects too

I might be able to set something up with the InstanceAssistant browser extension. It would be nice to have it in the main UI, but this could help temporarily.

I’ll send a PM :)

This is amazing, thank you for reaching out to the dev!

This was on the list of ideas for InstanceAssistant, but I didn’t know where to start. It is great news that it was added to the original extension.

That t-shirt would be cool, but don’t have time to properly test our extension before then

I’m looking forward to December when I’ll have time to work on it. I’ll try to attend some office hours, I didn’t know that was an option

Thank you for the suggestion! I’ll see about implementing this in a future version, unless someone else gets to it first

here is the issue I made: https://github.com/cynber/lemmy-instance-assistant/issues/53

Please let me know if the issue is missing a part of your suggestion or if there was any misunderstanding

Yep I think the defederation point is the big one which causes the idea to break down. I’ll edit the post to better reflect my thoughts now

Thank you!

Hope it helps :)

Look, honestly: if you want Facebook ot Twitter, go back to them.

This post was to talk about the merits/drawbacks of a potential change, and the constructive comments on the post have been helpful for that. Some of the other ‘solutions’ that have been posted here feel even more antithetical to the idea of decentralization (ex. redirecting upvotes, having communities follow other communities) so I was looking for a compromise that would address some of the annoyances without making the site another centralized platform. The intent was to allow users to choose how they want to link cross posts together, rather than having the community (or an app/frontend) make the decision for them. We’ve also been seeing users naturally gravitate to a few instances/communities, so I was looking for ways to redirect some of that traffic back to lesser known spaces.

Regardless, I appreciate the comment. Reading the perspectives on this post helped me see how locking the post completely would cause more issues and annoyances than it would help with. A simple “we are discussing X over on this post, feel free to join” seems like the better compromise.

that’s fair :)

Everyone who’s subscribed to the same communities will see all of each others’ comments.

This still relies on everyone using the same app/front-end.

I guess I’m thinking about how it would be helpful in more general cases. If someone has an issue with a FOSS app, and they ask about it in two small communities, it would be much better to have the troubleshooting discussion in one place rather than have both communities missing part of the context.

Ultimately in your example, the user can still make both posts, this doesn’t change that. It just directs the comments to one post’s comment section rather than having it spread out.

Still it’s good to think about cases where OP tries to abuse the system. Would a good middle ground just be the first implementation then? For OP to link to the post that they want to be the main discussion thread, but people are free to ignore that if they want.

I agree, instance assistant (our extension actually!) handles this, but it would be much better to have it exist natively.

this proposal would give more controls to the bigger communities

what are the concerns that come to mind?

I’d rather just have different communities

All the individual communities would continue to exist as they do right now. When it makes sense for a post to have a deeper discussion, users can lock and redirect accordingly

This works for viewing all the comments so far, but it doesn’t solve the discussion aspect since commentors from each community won’t be seeing or responding to the other comments. This is a bigger issue with smaller communities, where they’d mostly be top level comments / chains with minimal depth from each smaller community. Yes you can see all the comments, but the discussion quality is poor.

It’s also not as helpful when the automation fails. Something I’ve found is that the ‘crosspost’ field starts to get crowded on posts that link to a popular website. Combining comment sections from ALL of those posts isn’t as useful as having some intentional action from the OP.

A key aspect about this proposal is that it requires the OP to do something. If it doesn’t make sense for a community (ex. different intents behind the Politics communities), then OP shouldn’t lock their post. If OP does it anyway, then you can downvote that post.

I’ve been working on the !instance_assistant@lemmy.ca extension :)

RES initially helped me get into Reddit, so hopefully this helps people get into Lemmy as well

That’s a great point, and it was easy enough to add because it was very similar functionality. I just made a commit here.

I’ll upload version 1.2.6 tomorrow, and it should be approved and on the stores within a week or so. Thanks for the suggestion!

For smaller extensions, you can start by looking at the manifest.json as this file is the entry point. You can see which part of the code needs access to ‘everything’ and go from there.

https://developer.mozilla.org/en-US/docs/Mozilla/Add-ons/WebExtensions/manifest.json

For example, with Instance Assistant for Lemmy, you need to give access to all sites because it needs to run on any Lemmy site and that’s not something you can hardcode in.

This is very helpful, thank you! I’ll look into setting up more of that sometime

Good to know, I’ll explore some more

Thank you!

Makes sense, thank you!

Sounds good! This was my first dive into browser extensions as well. It’s not too bad once you go over the basics. If you give it a try, see the contributing page on the repo’s wiki for some resources on how to get started with browser extensions.

A super short summary is:

• manifest.json is the entry point, it links to HTML files (which represent things like the popup, sidebar) and scripts (which do most of the work)
• the background script runs all the time (see background.js), and the content scripts run on specific pages (ex. There’s one for Lemmy community pages, one for error pages)

If you DO give it a try, we were part way through migrating features from the LemmyTools userscript and that might be a good place to start. I wasn’t familiar with userscripts so I didn’t make much progress, and can’t get back to it for a little while. The issues page of the repo should have LemmyTools related features tagged. If any details are missing, let me know and I’ll add them in :)